Many of the investigations in this genetic epidemiology project arise from observations in families at high risk of cancer or in other etiologic studies. A pilot study to follow-up patients from a case-control study of 718 non-Hispanic white patients with cutaneous melanoma from melanoma clinics in Philadelphia and San Francisco is being planned. Questionnaire data, tumor, and DNA from a case-control study of 183 incident melanoma cases and 179 controls conducted in North-Eastern Italy were used to evaluate the melanocortin-1 receptor (MC1R) gene, the major regulator of human pigmentation, and somatic mutations of the BRAF oncogene. In subjects with melanoma arising on sun exposed areas of the body and with limited chronic solar damage, we found a strong association between MC1R germline variants and melanoma with somatic mutations in the BRAF oncogene. We confirmed this association in an independent population. We are now gathering data from other Italian populations to extend the analyses of association between melanoma risk and pigmentation and immune-related genes. We examined the major high-risk melanoma susceptibility gene, CDKN2A, in 703 registry-ascertained melanoma cases and 691 population-based Icelandic controls. The population-based prevalence of germline CDKN2A mutations was estimated to be 2% in invasive cutaneous melanoma patients. G89D was the major high penetrance CDKN2A mutation; relatives of patients with G89D showed increased cross-risks of pancreatic and head and neck cancers. The frequency of CDKN2A mutations in melanoma patients observed in this study was consistent with the only previous population-based estimates from predominantly North American and Australian populations. One of the potential pathways that leads to the development of melanoma includes the loss of regulation of common melanocytic nevi, which acquire atypic or dysplastic characteristics that can further evolve in neoplasia. To study this pathway, we are currently collecting multiple tissue samples of normal skin, common melanocytic nevi, dysplastic nevi, melanoma and metastasis from melanoma from the same subjects from North-Eastern Italy. We have extended the collection to another Italian site to increase the statistical power for analyses. When the sample size reaches 200 subjects, we plan to study the expression and presence of mutations in multiple genes of the cell cycle, telomere, and transduction pathways in the serial tissue samples and germline DNA to explore the mechanisms involved in melanoma development through nevi. Relatives of 365 of the glioma cases from a DCEG comprehensive case-control study of adults with brain tumors were interviewed about personal/family medical history and other risk factors. Analyses to examine the risk of cancer among the first degree relatives compared to population controls have been completed. Examination of questionnaire data from a retrospective single institution study of childhood medulloblastoma showed little evidence for increased cancer risks among relatives of medulloblastoma patients. Chordoma is a rare primary malignant bone tumor that arises mainly in the axial skeleton from rests of embryonic notochordal stem cells that failed to undergo normal regression. We are collaborating with the Department of Radiation Oncology, Massachusetts General Hospital, Boston, to identify chordoma families to help map and identify a chordoma susceptibility gene. The project involves obtaining personal and family medical history, buccal cells and slides of tumor tissue from the subgroup of patients most likely to have a genetic predisposition to chordoma: those diagnosed with chordoma less than 18 years. During the past year we completed enrollment of 54 chordoma patients. We are in the process of obtaining personal and family medical history, buccal cells and tumor tissue from them. In collaboration with Yale University Tissue Microarray (TMA) core facility, we have successfully built TMAs of 1,500 invasive tumors collected from the Polish Breast Cancer Study. We have immunohistochemically (IHC) stained a subset of these tumors (N=842) for 18 molecular markers involved in hormone biosynthesis, metabolism, and receptor mediated pathways. Analyses of these markers suggest that risk factors for breast cancer may vary by molecular subtypes and by hormone pathways characterized by co-expression of the hormonal markers. We have also stained all TMAs for six markers (ER-alpha, ER-beta, PR, HER2, EGFR, and CK5) using a newly developed Automated Quantitative Analysis (AQUA"). Comparison of the two approaches (AQUA and IHC) demonstrated AQUA analyses of tumors represented in TMAs provide reliable, quantitative measures of marker expression. Currently, we are evaluating some commercially available imaging analysis software to quantitatively measure IHC data in a high-throughput fashion. We are also exploring the utility of the imaging analysis in scoring marker expression in 1,547 non-invasive tissue cores including normal TDLUs and DCIS constructed on 32 TMA blocks. In addition to TMA analyses of candidate markers in fixed tissues, we are exploring an integrated approach that allows combined analyses of mRNA expression, copy number variation, microRNA, and gene methylation profiles to identify novel classifiers in frozen tumors from a subset of Polish cases. Renal cell carcinoma (RCC) rates in Central and Eastern Europe are among the highest in the world. The von Hippel Lindau-Hypoxia Inducible Factor (VHL-HIF) pathway has been implicated in kidney cancer tumorigenesis. We investigated the role of common variants in genes in the VHL"HIF pathway in the susceptibility of sporadic RCC and clear cell RCC. We identified common genetic variants in genes in the VHL-HIF pathway associated with kidney cancer risk. We are evaluating this finding in additional study populations. We are continuing to conduct studies based on new Swedish linked registry data expanded to include cohorts of cases of monoclonal gammopathy of uncertain significance (MGUS) and more complete ascertainment of Waldenstrom"s macroglobulinemia (WM) and lymphoplasmacytic lymphoma (LPL) cases. We found that the risk of lymphoproliferative (LP) malignancies were similar in relatives of WM versus LPL cases so we have combined these two entities. Relatives of WM/LPL patients were at increased risk of MGUS, WM/LPL, CLL, and other non-Hodgkin lymphomas (NHL) but not for myeloma (MM) or Hodgkin lymphoma (HL). Relatives of CLL cases were at increased risk for CLL, WM/LPL, and other NHLs but not for MM, HL, or MGUS. These familial patterns suggest that there are shared genes associated with susceptibility to multiple LP malignancies. The precursor, MGUS, is also included in this spectrum especially in relation to MM and WM/LPL. We are continuing to explore the role of autoimmune and other medical conditions on susceptibility to these tumors. We showed that the risk of NHL associated with celiac disease has decreased over time, which is probably associated with milder cases of celiac disease being diagnosed. We found that celiac disease in siblings aggregated with NHL suggesting common genes affecting both conditions. A new expanded data set containing outpatient registry diagnoses will enhance the power of these studies. We also have analyzed survival of NHL, HL, and CLL in relation to family history of LP malignancies. With the exception of T-cell/anaplastic lymphoma, survival patterns for patients with CLL, HL, and NHL with a family history of lymphoma were similar to those for sporadic patients, suggesting that most familial lymphomas do not have an altered clinical course.